stressapp.py

import numpy as np
import pickle
import streamlit as st
loaded_model=pickle.load(open('C:\\Users\\Tanya Gupta\\Downloads\\Iot_domain_project\\trained_model.sav','rb'))

#creating a function

def stress_prediction(input_data):
    

    # changing the input_data to numpy array
    input_data_as_numpy_array = np.asarray(input_data)
    input_data_as_numpy_array = input_data_as_numpy_array.astype(np.float64)
    # reshape the array as we are predicting for one instance
    input_data_reshaped = [input_data_as_numpy_array]

    prediction = loaded_model.predict(input_data_reshaped)
    print(prediction)

    if (prediction[0] == 0):
      return 'The person is not stressed'
    elif (prediction[0] == 1):
       return 'The person is slightly stressed'
    elif (prediction[0] == 2):
       return 'The person is stressed'
    elif(prediction[0]==3):
      return 'The person is highly stressed'
  
    
def main():
 
    # giving a title
    st.title('Stress level predictor')
    
    
    # getting the input data from the user
    
    body_temperature = st.text_input('Body temperature')
    limb_movement = st.text_input('Limb movement')
    Blood_oxygen = st.text_input('Blood oxygen')
    Sleeping_hours = st.text_input('Sleeping hours')
    Heart_rate = st.text_input('Heart rate')
    
    
    # code for Prediction
    diagnosis = ''
    
    # creating a button for Prediction
    
    if st.button('Predict stress'):
        diagnosis = stress_prediction([body_temperature, limb_movement, Blood_oxygen, Sleeping_hours, Heart_rate])
        
        
    st.success(diagnosis)
    
    
if __name__ == '__main__':
    main()